1. Field of the Invention
This invention relates generally to semiconductor devices. More particularly, the invention pertains to materials with high dielectric constants and methods for incorporating them in semiconductor devices.
2. State of the Art
In the manufacture and use of integrated circuit (IC) devices, new applications continually drive the development of devices with enhanced miniaturization and increased circuit density. Current and future developments in reducing the size of dynamic random access memory (DRAM) devices and the like, result in a need for storage capacitor materials having higher dielectric constants than currently available.
Capacitor cells are generally formed as “stacked” capacitors, i.e., positioned above the working surface of the chip or wafer, or “trench” capacitors, which are formed in a trench in the wafer or chip substrate. Because of the need to make the best use of available space in a device, current capacitor designs include nonplanar structures that may be formed in various configurations. References that describe examples of nonplanar capacitor constructions include U.S. Pat. No. 5,981,333 to Parekh et al., U.S. Pat. No. 5,981,350 to Geusic et al., U.S. Pat. No. 5,985,714 to Sandhu et al., and U.S. Pat. No. 5,985,732 to Fazan et al., each of which is incorporated herein by reference.
The number of high dielectric materials from which capacitor cells may be satisfactorily formed is limited. Insulating inorganic metal oxide materials such as ferroelectric or perovskite material have high dielectric constants and generally low leakage current. However, these materials require a step of “oxidation-densification” to produce the desired dielectric capacitor layer. Unfortunately, such oxidation-densification undesirably oxidizes the underlying electrode of conductively doped polysilicon. As practiced currently, an intervening oxygen-barrier layer is placed between the electrode and dielectric material. The barrier layer must be electrically conductive, inasmuch as the underlying polysilicon must be in electrical connection with the dielectric layer. The materials that may be used as oxygen barrier layers are limited in number. Elemental platinum on polysilicon has been suggested as a barrier layer for a lower capacitor plate but undergoes physical degradation with thermal cycling due to silicon diffusion through the platinum. Sputtered TiN and CVD-applied TiN have been known to fail due to diffusion along grain boundaries.
As known in the art, an alloy of titanium and tungsten may be used as a barrier layer between a silicon layer and an aluminum ohmic contact, where the junction is very shallow, i.e., less than about 0.2 μm.
In U.S. Pat. No. 5,985,714 having patentees of Sandhu et al. and of even assignment with this application, a condenser construction is described that uses a wide variety of dielectric materials including titanates of barium; barium and strontium; strontium; lead; barium and lead; lead and zirconium; lead and lanthanum; lead and lanthanum and zirconium; and bismuth. Lithium tantalate is also mentioned.
Several materials that have been used or undergone evaluation include Ta2O5 and (Ba,Sr)TiO3, the latter commonly known as BST. Ta2O5 has a dielectric constant k that is about 15 to 25; the dielectric constant is too low to meet the requirements for use in advanced DRAM and other semiconductor construction, i.e., a much higher dielectric constant generally exceeding about 100.
BST materials have dielectric constants, i.e., about 300 to 600, which are higher than dielectric materials in current use. However, the processes for producing BST are not yet fully developed. The processing of BST is intrinsically difficult because of the low volatility of the precursors used in the chemical vapor deposition (CVD) step, and by difficulty in controlling the complex stoichiometry to maintain the desired material characteristics.
Alternative dielectric materials have appeared to offer potential advantages in dielectric constant value and ease of manufacture. For example, TiO2 films are well known as high dielectric materials. TiO2 films have a dielectric constant greater than 100, which is considerably higher than that of Ta2O5. In addition, TiO2 films may be formed using current manufacturing methods. However, it has been found that capacitors made of pure TiO2 have a high leakage current unacceptable in high-density devices required by current and developing electronic technology.
It has been shown by Matsushita (Jpn. J. Appl. Phys. 30 (1991) 3594) that doping TiO2 with SiO2 may dramatically improve the leakage current of the TiO2 materials used in capacitors. However, this doped material is generally comparable to Ta2O5 in dielectric constant, i.e., in a low range of about 15 to 25.
Other materials considered for high dielectric use include tungsten trioxide (WO3) but it has an unacceptably high leakage current.
Commercial production of semiconductor devices requires a sequence of basic physical/chemical processes, many of which are typically performed on a large number of dice in a semiconductor wafer prior to singulating and packaging the devices. The minimal time required to carry out the process from beginning to end is extensive, with high attendant cost. For example, it usually takes about six to eight weeks or more to produce a potentially finished memory chip from an uncut multi-wafer crystal. It is desirable to shorten the processing time as much as possible, to reduce manpower cost and increase the throughput rate of processing equipment.
The instant invention addresses the need for new dielectric materials having high dielectric constants (K) of about 100 or more, and the capability of being processed more quickly, easily and precisely, and at a lower cost than other high dielectric material candidates.